Separate neural representations of prediction error valence and surprise

Evidence from an fMRI meta‐analysis

Elsa Fouragnan, Chris Retzler, Marios G. Philiastides

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Learning occurs when an outcome differs from expectations, generating a reward prediction error signal (RPE). The RPE signal has been hypothesized to simultaneously embody the valence of an outcome (better or worse than expected) and its surprise (how far from expectations). Nonetheless, growing evidence suggests that separate representations of the two RPE components exist in the human brain. Meta‐analyses provide an opportunity to test this hypothesis and directly probe the extent to which the valence and surprise of the error signal are encoded in separate or overlapping networks. We carried out several meta‐analyses on a large set of fMRI studies investigating the neural basis of RPE, locked at decision outcome. We identified two valence learning systems by pooling studies searching for differential neural activity in response to categorical positive‐versus‐negative outcomes. The first valence network (negative > positive) involved areas regulating alertness and switching behaviours such as the midcingulate cortex, the thalamus and the dorsolateral prefrontal cortex whereas the second valence network (positive > negative) encompassed regions of the human reward circuitry such as the ventral striatum and the ventromedial prefrontal cortex. We also found evidence of a largely distinct surprise‐encoding network including the anterior cingulate cortex, anterior insula and dorsal striatum. Together with recent animal and electrophysiological evidence this meta‐analysis points to a sequential and distributed encoding of different components of the RPE signal, with potentially distinct functional roles.
Original languageEnglish
Pages (from-to)2887-2906
Number of pages20
JournalHuman Brain Mapping
Volume39
Issue number7
Early online date25 Mar 2018
DOIs
Publication statusPublished - 1 Jul 2018

Fingerprint

Reward
Magnetic Resonance Imaging
Prefrontal Cortex
Learning
Gyrus Cinguli
Thalamus
Brain

Cite this

@article{5a52db15601e4bbf9428b0267b6900fd,
title = "Separate neural representations of prediction error valence and surprise: Evidence from an fMRI meta‐analysis",
abstract = "Learning occurs when an outcome differs from expectations, generating a reward prediction error signal (RPE). The RPE signal has been hypothesized to simultaneously embody the valence of an outcome (better or worse than expected) and its surprise (how far from expectations). Nonetheless, growing evidence suggests that separate representations of the two RPE components exist in the human brain. Meta‐analyses provide an opportunity to test this hypothesis and directly probe the extent to which the valence and surprise of the error signal are encoded in separate or overlapping networks. We carried out several meta‐analyses on a large set of fMRI studies investigating the neural basis of RPE, locked at decision outcome. We identified two valence learning systems by pooling studies searching for differential neural activity in response to categorical positive‐versus‐negative outcomes. The first valence network (negative > positive) involved areas regulating alertness and switching behaviours such as the midcingulate cortex, the thalamus and the dorsolateral prefrontal cortex whereas the second valence network (positive > negative) encompassed regions of the human reward circuitry such as the ventral striatum and the ventromedial prefrontal cortex. We also found evidence of a largely distinct surprise‐encoding network including the anterior cingulate cortex, anterior insula and dorsal striatum. Together with recent animal and electrophysiological evidence this meta‐analysis points to a sequential and distributed encoding of different components of the RPE signal, with potentially distinct functional roles.",
keywords = "decision making, learning, meta-analysis, prediction error, reward",
author = "Elsa Fouragnan and Chris Retzler and Philiastides, {Marios G.}",
note = "This is the peer reviewed version of the following article: Fouragnan E, Retzler C, PhiliastidesMG. Separate neural representations of prediction error valenceand surprise: Evidence from an fMRI meta-analysis. Hum BrainMapp. 2018;00:1–20.https://doi.org/10.1002/hbm.24047, which has been published in final form at https://onlinelibrary.wiley.com/doi/epdf/10.1002/hbm.24047. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving.",
year = "2018",
month = "7",
day = "1",
doi = "10.1002/hbm.24047",
language = "English",
volume = "39",
pages = "2887--2906",
journal = "Human Brain Mapping",
issn = "1065-9471",
publisher = "Wiley-Liss Inc.",
number = "7",

}

Separate neural representations of prediction error valence and surprise : Evidence from an fMRI meta‐analysis. / Fouragnan, Elsa; Retzler, Chris; Philiastides, Marios G.

In: Human Brain Mapping, Vol. 39, No. 7, 01.07.2018, p. 2887-2906.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Separate neural representations of prediction error valence and surprise

T2 - Evidence from an fMRI meta‐analysis

AU - Fouragnan, Elsa

AU - Retzler, Chris

AU - Philiastides, Marios G.

N1 - This is the peer reviewed version of the following article: Fouragnan E, Retzler C, PhiliastidesMG. Separate neural representations of prediction error valenceand surprise: Evidence from an fMRI meta-analysis. Hum BrainMapp. 2018;00:1–20.https://doi.org/10.1002/hbm.24047, which has been published in final form at https://onlinelibrary.wiley.com/doi/epdf/10.1002/hbm.24047. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving.

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Learning occurs when an outcome differs from expectations, generating a reward prediction error signal (RPE). The RPE signal has been hypothesized to simultaneously embody the valence of an outcome (better or worse than expected) and its surprise (how far from expectations). Nonetheless, growing evidence suggests that separate representations of the two RPE components exist in the human brain. Meta‐analyses provide an opportunity to test this hypothesis and directly probe the extent to which the valence and surprise of the error signal are encoded in separate or overlapping networks. We carried out several meta‐analyses on a large set of fMRI studies investigating the neural basis of RPE, locked at decision outcome. We identified two valence learning systems by pooling studies searching for differential neural activity in response to categorical positive‐versus‐negative outcomes. The first valence network (negative > positive) involved areas regulating alertness and switching behaviours such as the midcingulate cortex, the thalamus and the dorsolateral prefrontal cortex whereas the second valence network (positive > negative) encompassed regions of the human reward circuitry such as the ventral striatum and the ventromedial prefrontal cortex. We also found evidence of a largely distinct surprise‐encoding network including the anterior cingulate cortex, anterior insula and dorsal striatum. Together with recent animal and electrophysiological evidence this meta‐analysis points to a sequential and distributed encoding of different components of the RPE signal, with potentially distinct functional roles.

AB - Learning occurs when an outcome differs from expectations, generating a reward prediction error signal (RPE). The RPE signal has been hypothesized to simultaneously embody the valence of an outcome (better or worse than expected) and its surprise (how far from expectations). Nonetheless, growing evidence suggests that separate representations of the two RPE components exist in the human brain. Meta‐analyses provide an opportunity to test this hypothesis and directly probe the extent to which the valence and surprise of the error signal are encoded in separate or overlapping networks. We carried out several meta‐analyses on a large set of fMRI studies investigating the neural basis of RPE, locked at decision outcome. We identified two valence learning systems by pooling studies searching for differential neural activity in response to categorical positive‐versus‐negative outcomes. The first valence network (negative > positive) involved areas regulating alertness and switching behaviours such as the midcingulate cortex, the thalamus and the dorsolateral prefrontal cortex whereas the second valence network (positive > negative) encompassed regions of the human reward circuitry such as the ventral striatum and the ventromedial prefrontal cortex. We also found evidence of a largely distinct surprise‐encoding network including the anterior cingulate cortex, anterior insula and dorsal striatum. Together with recent animal and electrophysiological evidence this meta‐analysis points to a sequential and distributed encoding of different components of the RPE signal, with potentially distinct functional roles.

KW - decision making

KW - learning

KW - meta-analysis

KW - prediction error

KW - reward

UR - http://www.scopus.com/inward/record.url?scp=85044410810&partnerID=8YFLogxK

U2 - 10.1002/hbm.24047

DO - 10.1002/hbm.24047

M3 - Article

VL - 39

SP - 2887

EP - 2906

JO - Human Brain Mapping

JF - Human Brain Mapping

SN - 1065-9471

IS - 7

ER -